Hybrid Approach for Underwater Image Restoration and Enhancement
Gary Sequeira, Vidyasagar Mekkalki, Jeevan Prabhu, Samarth Borkar, Mangish Desai · 2021 International Conference on Emerging Smart Computing and Informatics (ESCI) · 2021
Application of image processing algorithms to underwater images results in unrealistic pictures with non-uniform scene lighting. In this paper, we propound an innovative algorithm utilizing image restoration and image enhancement techniques, into a single algorithm for underwater image processing. Our image restoration algorithm is a modified approach based on automatic red channel algorithm for the blue channel, since the red color channel does not always have the least intensity underwater. To refine the contrast of the image, we use a modified Integrated Color Model for image enhancement. Results of the proposed algorithm show realistic, high contrast and clear underwater images as compared to other algorithms for the same set of images. To validate our enhanced results, we have used entropy and histogram analysis as a metric for quantitative analysis.